Collaborative Research: PrimaryAI: Integrating Artificial Intelligence into Upper Elementary Science with Immersive Problem-Based Learning

人工智能已经成为一项基础技术，正在深刻地重塑社会。随着人工智能和机器学习能力的快速发展，这些技术正迅速在经济的各个领域得到广泛应用。越来越多的人认识到对人工智能劳动力的需求，这凸显了迫切需要深入了解如何向K-12学生介绍人工智能，以及如何支持K-12教师在这方面的努力。因为小学阶段是学生发展的关键时期？对STEM的积极看法和倾向，为小学生创造引人入胜的人工智能学习体验是至关重要的。同样，在小学阶段发展生命科学的学科核心思想对于培养不同学习者对STEM的持久理解和兴趣非常重要。然而，人工智能在基础教育中明显缺席，关于人工智能在基础教育中的学习和教学的研究也很有限。人工智能基础教育的一个关键开放问题是，如何在应用人工智能解决核心科学问题的背景下，向学生介绍人工智能的基础知识？这个问题带来了重大挑战，因为解决这个问题需要对学生的学习过程和结果进行社会认知解释，这可以用来为人工智能和科学综合课程的设计提供信息。通过将人工智能嵌入基础生命科学教育，该项目的研究人员将研究如何满足有针对性的人工智能教育的需求，同时为基础阶段的生命科学学习创造创新方法。本项目由STEM+ Computing （STEM+C）项目资助，该项目支持研究和开发，以了解STEM学习中计算和计算思维的整合。该项目将解决三个研究问题：1)如何利用沉浸式问题学习为小学高年级学生创造融合人工智能和生命科学的引人入胜的学习体验？2)如何在小学高年级课堂设计一个人工智能与生命科学相结合的教师专业发展模式？3)沉浸式基于问题的学习在哪些方面支持小学生学习人工智能和生命科学？为了解决第一个研究问题，项目团队将迭代地设计、开发和完善PrimaryAI，这是一个集成的人工智能科学课程和沉浸式学习环境，将人工智能概念包括感知、规划、机器人、机器学习以及人工智能伦理引入小学高年级科学课堂。PrimaryAI将使学生能够通过使用适合年龄的人工智能工具来解决科学冒险中的生态问题，从而协同学习人工智能，因为他们参与证据论证，分析和解释数据，开发模型并构建解释。为了解决第二个研究问题，项目团队将创建PrimaryAI专业发展模型。该模型将帮助教师在课堂上忠实地使用PrimaryAI。它将采取实践社区的形式，围绕三个关键要素设计：教师专业学习，指导和在线社区。教师学习将以指导和参与式共同设计的沉浸式问题学习环境为中心，确保教师深入了解人工智能注入的生命科学教育。为了解决第三个研究问题，项目团队将进行基于设计的研究，以调查PrimaryAI如何提高学生对以AI概念为中心的计算的理解，以及对学科生命科学内容和实践的理解。学生的学习和参与将通过以下方式进行评估：1)视频分析和互动分析；2)焦点小组，包括专题分析；3)对学生进行访谈，以试点原型和措施；4)实施的跨案例分析，包括学生参与标题编码；5)对人工智能和生命科学内容的职前测量。为了评估专业发展模式，教师课堂实践将通过以下方式进行衡量：1)视频分析和共同设计与实施的互动分析；2)对教师教案、期刊、材料、笔记和反思的分析，包括保真度和适应性、参与编码、启发式案例研究和互动分析。该项目的可交付成果将包括PrimaryAI课程、PrimaryAI沉浸式问题学习环境、PrimaryAI专业发展模型及其相关材料，以及PrimaryAI在线社区门户网站。该项目的成果将为小学高年级人工智能生命科学学习环境和教学模式的设计和开发提供知识。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Artificial intelligence has emerged as a foundational technology that is profoundly reshaping society. With rapid advances in a wide array of AI and machine learning capabilities, these technologies are quickly finding broad application in every sector of the economy. The growing recognition of the demand for an AI-literate workforce highlights the urgent need to develop a deep understanding of how to introduce K-12 students to AI and how to support K-12 teachers in this endeavor. Because the elementary grades are a critical time for developing students? positive perceptions and dispositions toward STEM, creating engaging AI learning experiences for elementary grade students is of paramount importance. Similarly, developing disciplinary core ideas in life science in the elementary grades is important for creating enduring understanding of and interest in STEM for diverse learners. However, AI has been conspicuously absent from elementary education, and there has been limited research examining AI learning and teaching at the elementary level. A key open question for AI elementary education is how can students be introduced to the fundamentals of AI in the context of its application to solving core science problems? This question poses significant challenges because addressing it entails developing a socio-cognitive account of student learning processes and outcomes that can be used to inform the design of an integrated AI and science curriculum. By embedding AI in elementary life science education, researchers of this project will investigate how to meet the demand for targeted AI education while simultaneously creating innovative approaches to robust life science learning at the elementary level. This project is funded by the STEM + Computing (STEM+C) program that supports research and development to understand the integration of computing and computational thinking in STEM learning.The project will address three research questions: 1) How can we create engaging learning experiences integrating artificial intelligence and life science for upper elementary students by leveraging immersive problem-based learning? 2) How can we design a teacher professional development model for integrating artificial intelligence and life science in upper elementary classrooms? and 3) In what ways does engagement with immersive problem-based learning support upper elementary students' learning artificial intelligence and life science? To address the first research question, the project team will iteratively design, develop, and refine PrimaryAI, an integrated AI-science curriculum and immersive learning environment that will introduce AI concepts including perception, planning, robotics, and machine learning, as well as AI ethics, into upper elementary science classrooms. PrimaryAI will enable students to collaboratively learn about artificial intelligence by using age-appropriate AI tools to solve ecology problems in science adventures as they engage in argument from evidence, analyze and interpret data, develop models, and construct explanations. To address the second research question, the project team will create the PrimaryAI professional development model. The model will prepare teachers to use PrimaryAI with fidelity within their classrooms. It will take the form of a community of practice designed around three key elements: teacher professional learning, coaching, and an online community. Teacher learning will center on mentoring and participatory co-design of the immersive problem-based learning environment to ensure deep teacher knowledge of AI-infused life science education. To address the third research question, the project team will conduct design-based research to investigate how PrimaryAI improves students' understanding of computing centered around AI concepts, and of disciplinary life science content and practices. Student learning and engagement will be assessed using 1) video analysis and interaction analysis, 2) focus groups, including thematic analyses, 3) interviews with students to pilot prototypes and measures, 4) cross-case analyses of implementations, including student engagement rubric coding, 5) pre-post measures on artificial intelligence and life science content. To assess the professional development model, teacher classroom practice will be measured with 1) video analysis and interaction analysis of co-design and implementation, and 2) analyses of teacher lesson plans, journals, materials, notes, and reflections, including fidelity and adaptation, engagement coding, heuristic case studies, and interaction analyses. The deliverables of the project will include the PrimaryAI curricula, the PrimaryAI immersive problem-based learning environment, the PrimaryAI professional development model and its associated materials, and the PrimaryAI online community portal. The outcome of this project will build knowledge on the design and development of AI-infused life science learning environments and teaching models for upper elementary grades.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.